CN109099964A - Mechanical seal end surface state monitoring method - Google Patents
Mechanical seal end surface state monitoring method Download PDFInfo
- Publication number
- CN109099964A CN109099964A CN201810754424.5A CN201810754424A CN109099964A CN 109099964 A CN109099964 A CN 109099964A CN 201810754424 A CN201810754424 A CN 201810754424A CN 109099964 A CN109099964 A CN 109099964A
- Authority
- CN
- China
- Prior art keywords
- face
- mechanical seal
- end surface
- layer
- texturing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Abstract
The invention discloses mechanical seal end surface state monitoring method, method includes the following steps: step S1: being sealed to machinery;Step S2: the pattern of seal face is determined;Step S3: texturing processing is carried out according to surface of the pattern of seal face to mechanical seal end surface;Step S4: the texturing parameter under selection mechanical seal operation demand optimum state is utilized respectively DLC coating end face and the compound same seal face of texturing end face processing;Step S5: it carries out Experimental Comparison and analyzes the frictional behavior between untreated end face, DLC coating end face and compound texturing end face;Solve existing sealing premature failure, caused by downtime is long, enterprises' loss is big problem.
Description
Technical field
The present invention relates to mechanical seal end surface detection fields, especially mechanical seal end surface state monitoring method.
Background technique
Rotating machinery, such as: compressor, steam turbine, gas turbine, the hydraulic turbine, centrifugal pump are modern industrial production processes
In key equipment.The processing medium of such mechanical equipment processing usually has inflammable, explosive, toxic, high temperature and corrosivity.For
It prevents or limitation processing medium is leaked in natural environment along the shaft end of rotating machinery, it is necessary to use effective and reliable shaft end
Mechanical seal (Mechanical Seal) device, while must also consider that selected axle end sealing device is able to satisfy modern work
Industry is environmentally protective, pollution-free, long-life, high benefit and the requirement for completing energy-saving and emission-reduction index etc..According to English, beauty, De Deng state
Statistics, in petrochemical industry equipment, centrifugation pump accounts for 85% or more of whole pump equipment, and is centrifuged pump maintenance mans
Work measure 70% or so be solve the problems, such as sealing cause.
Such as in the entire energy consumption of China's petrochemical industry, the large percentage that various fluid dynamic machinery occupy, these machines
The interior leakage or leakage of tool often result in volumetric efficiency reduction, and power loss reaches 10% or more of general power, and small-sized machine is even
It is also possible to pollution environment simultaneously up to 40-50%.Novel mechanical seal form is studied, can not only reduce processing medium because of leakage
And the material loss generated, energy required for overcoming mechanical seal end surface to rub can also be reduced.Therefore, further investigation is mechanical
Sealing can ensure sealing element itself and corollary system reliability of operation and stability, reduce the consumption of material and the energy, always
It is an important topic in equipment of industrial product technology.
Compared with common fillers sealing, mechanical seal is with leakage rate is low, long service life, driving power consumption is low and does not wear
The features such as shaft.But because of complex structural designs, especially rubbing surface need to make micro-scale surface processing, and processing technology is more demanding etc.
Reason, mechanical seal product needed for leading to each relevant industries in China is by the whole world a few major company (such as Flowserve, John
Crane, Eagle Burgmann) it is monopolized.Meanwhile external mechanical seal product is compared, homemade goods high parameter field also
There are larger gaps.By taking dry gas seals product (a kind of non-contacting mechanical seal of surface of friction pair) as an example, the production of foreign countries' exploitation
Product are using pressure up to 45MPa, and linear velocity 230m/s, homemade goods maximum working (operation) pressure (MWP) is 10MPa, linear velocity 150m/s with
It is interior.
In order to improve domestic machinery sealing element operating parameter, the product of China's independent research is made to meet various application environments
Demand must just carry out further investigation in terms of mechanical seal mechanism, simulation calculation, status monitoring.In addition, domestic machinery seals
Although product can reach preferable effect in actual use, but often there is the case where sealing premature failure.
The reason of in order to find sealing premature failure, downtime is reduced, reduces enterprises' loss, mechanical seal mechanism must be just unfolded
Deeper into research.
In conclusion the mechanical seal market demand is big, have a extensive future.Improve China's mechanical seal independent research water
It is flat, reinforce mechanical seal theoretical research, to improving existing machinery Seal Design, shorten product development cycle, quickly flexibly ring
Answer user demand significant.Particularly, in more microcosmic point research seal face in system start and stop and steady operational process
Middle pressure field, thermal field, Flow Field Distribution carry out multi- scenarios method analysis to mechanical seal, will such as axial float, it is angular swing etc. it is dry
It disturbs factor and Accurate Analysis model is added, the above perspective study is carried out to mechanical seal mechanism, China's mechanical seal skill can be made
Art further shortens the gap between developed country, is preferably national economy service.
Summary of the invention
To solve problems of the prior art, the present invention provides mechanical seal end surface state monitoring methods, solve
Existing sealing premature failure, caused by downtime is long, enterprises' loss is big problem.
The technical solution adopted by the present invention is that mechanical seal end surface state monitoring method, method include the following steps:
Mechanical seal end surface state monitoring method, method include the following steps:
Step S1: machinery is sealed;
Step S2: the pattern of seal face is determined;
Step S3: texturing processing is carried out according to surface of the pattern of seal face to mechanical seal end surface;
Step S4: the texturing parameter under selection mechanical seal operation demand optimum state is utilized respectively DLC coating end face
With the compound same seal face of texturing end face processing;
Step S5: it carries out Experimental Comparison and analyzes rubbing between untreated end face, DLC coating end face and compound texturing end face
Wipe characteristic;
Step S6: analyzing the Experimental Comparison result of three kinds of obtained frictional behaviors of step S5, and is tied according to analysis
Fruit carries out signal data acquisition;
Step S7: using blind source separating and variation mode decomposition method to the signal extraction validity feature value of acquisition;
Step S8: according to the validity feature value extracted, using RBF network model and B-spline network model to abrasion shape
State is identified;
Step S9: output recognition result completes the monitoring to mechanical seal end surface state, terminates program.
Mechanical seal end surface state monitoring method of the present invention has the beneficial effect that:
1. screening is for the texturing parameter and DLC coating layer thickness under mechanical seal operation demand optimum state, by two kinds
Processing mode combines, and handles same seal face, carries out Experimental Comparison later and analyzes untreated end face, texturing end face, DLC
Frictional behavior between coating end face and compound texturing DLC coating end face.
2. the signal processing methods such as blind source separating and variation mode decomposition are introduced and extract effectively spy by the signal of pair acquisition
Sign;On the basis of studying vibration information, acoustic emission information and ultrasound information, by establishing B-spline neural network to close
Envelope state is monitored, and studying has distinct novelty.
Detailed description of the invention
Fig. 1 is the general flow chart of mechanical seal end surface state monitoring method of the present invention.
Fig. 2 is the RBF network architecture schematic diagram of mechanical seal end surface state monitoring method of the present invention.
Fig. 3 is the B-spline neural network structure schematic diagram of mechanical seal end surface state monitoring method of the present invention.
Specific embodiment
A specific embodiment of the invention is described below, in order to facilitate understanding by those skilled in the art this hair
It is bright, it should be apparent that the present invention is not limited to the ranges of specific embodiment, for those skilled in the art,
As long as various change is in the spirit and scope of the present invention that the attached claims limit and determine, these variations are aobvious and easy
See, all are using the innovation and creation of present inventive concept in the column of protection.
As shown in Figure 1, mechanical seal end surface state monitoring method, method include the following steps:
Step S1: machinery is sealed;
Step S2: the pattern of seal face is determined;
Step S3: texturing processing is carried out according to surface of the pattern of seal face to mechanical seal end surface;
Step S4: the texturing parameter under selection mechanical seal operation demand optimum state is utilized respectively DLC coating end face
With the compound same seal face of texturing end face processing;
Step S5: it carries out Experimental Comparison and analyzes rubbing between untreated end face, DLC coating end face and compound texturing end face
Wipe characteristic;
Step S6: analyzing the Experimental Comparison result of three kinds of obtained frictional behaviors of step S5, and is tied according to analysis
Fruit carries out signal data acquisition;
Step S7: using blind source separating and variation mode decomposition method to the signal extraction validity feature value of acquisition;
Step S8: according to the validity feature value extracted, using RBF network model and B-spline network model to abrasion shape
State is identified;
Step S9: output recognition result completes the monitoring to mechanical seal end surface state, terminates program.
The result of the step S8 of this programme is as follows:
RBF neural is 93.8% in the discrimination of pressure medium 0.2MPa, the identification under 0.5MPa pressure medium
Rate is 90.6%, and the discrimination under 1MPa pressure medium is 87.5%;
Discrimination of the B-spline neural network in 0.2Mpa is 91.5%, and the discrimination under 0.5MPa pressure medium is
92.6%, the discrimination under 1MPa pressure medium is 93.8%.
The characteristic extraction step of the blind source separating of the step S7 of this programme is as follows:
Step A1: s=(s is set1,s2,…,sn)TIt is the independent source of n zero-mean, x=(x1,x2,…,xm)TIt is m
Observation signal, observation signal are formed by independent source linear hybrid, and A is mxn rank full rank hybrid matrix, are not considering that additivity is made an uproar
In the case where sound, ICA mixed model is represented by x=As;
Step A2: a separation matrix W is found, it is independent as far as possible through the transformed output y of W to make observation signal x, and be
The estimation of source signal:
Y=Wx.
The characteristic extraction step of the variation mode decomposition method of step S7 is as follows:
Step B1: original signal f is decomposed into K basic friction angle component, the expression formula corresponding to VMD model are as follows:
In formula, { uk}={ u1,…,uKIt is to decompose K obtained Intrinsic mode functions;{ωk}={ ω1,…,ωKTable
Show that the center frequency of each component, σ (t) are dirac (Dirac) function, * represents convolution, and f is original signals and associated noises.
Step B2: introducing secondary penalty factor a and Lagrange multiplier operator λ, by restrictive variational problem become it is non-about
Beam variational problem, wherein secondary penalty factor can guarantee the reconstruction accuracy of signal, glug in the presence of Gaussian noise
Bright day operator makes constraint condition keep stringency, to obtain the optimal solution of problem:
In formula, a is secondary penalty factor, and λ (t) is Lagrange multiplier operator.
The texturing parameter of the step S4 of this programme includes texture aspect ratio, texture features and texture density.
As shown in Fig. 2, the RBF network model expression formula of the step S8 of this programme are as follows:
In formula, W is connection weight vector, W=(ωk1,ωk2,…ωkm), Φ is hidden layer output,
As shown in figure 3, the B-spline network model of the step S8 of this programme is
Ith layer is input layer,
In formula, i=1,2 ..., n;N is input number.
IIth layer is blurring layer, using B-spline function as membership function,
In formula,It is the basic function of j-th of fuzzy set, i=1,2 ..., n;J=1,2 ..., h, h are B-spline basic function
Number;K is the order of B-spline function;
IIIth layer is rules layer,
In formula, ki=1,2 ... h, " * " are product operations, indicate minimizing operation;
IVth layer is output layer,
In formula, l=1,2 ..., t;T is that network exports number;ωk1k2…knIndicate the connection weight of network.
The present embodiment is when implementing, the first step, carries out texturing, DLC coating respectively to mechanical seal end surface, screens needle
To the texturing parameter and DLC coating layer thickness under mechanical seal operation demand optimum state, two kinds of processing modes are combined, are located
Same seal face is managed, Experimental Comparison is carried out later and analyzes between untreated end face, DLC coating end face and compound texturing end face
Frictional behavior;
The signal processing methods such as blind source separating and variation mode decomposition are introduced and are extracted to the signal of acquisition by second step
Validity feature;
Useful signal after blind source separating is extracted into following characteristics:
Root mean square (RMS):
Energy variance (EV):
Peak factor (PAR):
xpeakFor the peak value of signal
Centre frequency (CF):
piIt is signal in frequency fiWhen amplitude
Flexure (AS):
xmeanFor the average value of signal
After the processing of features described above parameter normalization, the results are shown in Table 1:
1 normalization characteristic of table
After VMD is decomposed, the percentage E of gross energy shared by preceding 5 IMF components is extractedi/ E, gross energy
The results are shown in Table 2.
Table 2
Third step passes through RBF network mould on the basis of studying vibration information, acoustic emission information and ultrasound information
Type and B-spline network model identify state of wear, obtain report result.
RBF network model expression formula are as follows:
In formula, W is connection weight vector, W=(ωk1,ωk2,…ωkm), Φ is hidden layer output,
B-spline network model is
Ith layer is input layer,
In formula, i=1,2 ..., n;N is input number.
IIth layer is blurring layer, using B-spline function as membership function,
In formula,It is the basic function of j-th of fuzzy set, i=1,2 ..., n;J=1,2 ..., h, h are B-spline basic function
Number;K is the order of B-spline function;
IIIth layer is rules layer,
In formula, ki=1,2 ... h, " * " are product operations, indicate minimizing operation;
IVth layer is output layer,
In formula, l=1,2 ..., t;T is that network exports number;ωk1k2…knIndicate the connection weight of network.
RBF neural is 93.8% in the discrimination of pressure medium 0.2MPa, the identification under 0.5MPa pressure medium
Rate is 90.6%, and the discrimination under 1MPa pressure medium is 87.5%.
Discrimination of the B-spline neural network in 0.2Mpa is 91.5%, and the discrimination under 0.5MPa pressure medium is
92.6%, the discrimination under 1MPa pressure medium is 93.8%.
Claims (6)
1. mechanical seal end surface state monitoring method, which is characterized in that method includes the following steps:
Step S1: machinery is sealed;
Step S2: the pattern of seal face is determined;
Step S3: texturing processing is carried out according to surface of the pattern of seal face to mechanical seal end surface;
Step S4: the texturing parameter under selection mechanical seal operation demand optimum state is utilized respectively DLC coating end face and answers
Close the same seal face of texturing end face processing;
Step S5: it carries out Experimental Comparison and analyzes the spy of the friction between untreated end face, DLC coating end face and compound texturing end face
Property;
Step S6: analyzing the Experimental Comparison result of three kinds of obtained frictional behaviors of step S5, and based on the analysis results into
Row signal data acquisition;
Step S7: using blind source separating and variation mode decomposition method to the signal extraction validity feature value of acquisition;
Step S8: according to the validity feature value extracted, using RBF network model and B-spline network model to state of wear into
The identification of row feature;
Step S9: output recognition result completes the monitoring to mechanical seal end surface state, terminates program.
2. mechanical seal end surface state monitoring method according to claim 1, which is characterized in that the blind source of the step S7
Isolated characteristic extraction step is as follows:
Step A1: s=(s is set1,s2,…,sn)TIt is the independent source of n zero-mean, x=(x1,x2,…,xm)TIt is observed for m
Signal, observation signal are formed by independent source linear hybrid, and A is m × n rank full rank hybrid matrix, are not considering additive noise
In the case where, ICA mixed model is represented by x=As;
Step A2: finding a separation matrix W, keep observation signal x independent as far as possible through the transformed output y of W, and believes for source
Number estimation:
Y=Wx.
3. mechanical seal end surface state monitoring method according to claim 1, which is characterized in that the variation of the step S7
The characteristic extraction step of mode decomposition method is as follows:
Step B1: original signal f is decomposed into K basic friction angle component, the expression formula corresponding to VMD model are as follows:
In formula, { uk}={ u1,…,uKIt is to decompose K obtained Intrinsic mode functions;{ωk}={ ω1,…,ωKIndicate each
The center frequency of a component, σ (t) are dirac (Dirac) function, and * represents convolution, and f is original signals and associated noises.
Step B2: introducing secondary penalty factor a and Lagrange multiplier operator λ, and restrictive variational problem is become non-binding
Variational problem, wherein secondary penalty factor can guarantee the reconstruction accuracy of signal in the presence of Gaussian noise, Lagrange
Operator makes constraint condition keep stringency, to obtain the optimal solution of problem:
In formula, a is secondary penalty factor, and λ (t) is Lagrange multiplier operator.
4. mechanical seal end surface state monitoring method according to claim 1, which is characterized in that the texture of the step S4
Changing parameter includes texture aspect ratio, texture features and texture density.
5. mechanical seal end surface state monitoring method according to claim 1, which is characterized in that the RBF of the step S8
Network model expression formula are as follows:
In formula, W is connection weight vector, W=(ωk1,ωk2,…ωkm), Φ is hidden layer output,
6. mechanical seal end surface state monitoring method according to claim 1, which is characterized in that the B sample of the step S8
Network model is
Ith layer is input layer,
In formula, i=1,2 ..., n;N is input number.
IIth layer is blurring layer, using B-spline function as membership function,
In formula,It is the basic function of j-th of fuzzy set, i=1,2 ..., n;J=1,2 ..., h, h are of B-spline basic function
Number;K is the order of B-spline function;
IIIth layer is rules layer,
In formula, ki=1,2 ... h, " * " are product operations, indicate minimizing operation;
IVth layer is output layer,
In formula, l=1,2 ..., t;T is that network exports number;ωk1k2…knIndicate the connection weight of network.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810754424.5A CN109099964A (en) | 2018-07-11 | 2018-07-11 | Mechanical seal end surface state monitoring method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810754424.5A CN109099964A (en) | 2018-07-11 | 2018-07-11 | Mechanical seal end surface state monitoring method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109099964A true CN109099964A (en) | 2018-12-28 |
Family
ID=64845928
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810754424.5A Pending CN109099964A (en) | 2018-07-11 | 2018-07-11 | Mechanical seal end surface state monitoring method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109099964A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114644276A (en) * | 2022-04-11 | 2022-06-21 | 伊萨电梯有限公司 | Intelligent elevator control method under mixed scene condition |
CN115754010A (en) * | 2022-11-10 | 2023-03-07 | 四川科力特硬质合金股份有限公司 | Quality control method for sealing end face of mechanical sealing ring |
CN116989937A (en) * | 2023-09-25 | 2023-11-03 | 苏州俊煌机械科技有限公司 | Detection method and device for mechanical sealing element |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102313578A (en) * | 2011-08-04 | 2012-01-11 | 广州市香港科大霍英东研究院 | Mechanical seal online monitoring system |
CN105115453A (en) * | 2015-04-27 | 2015-12-02 | 南京林业大学 | Device and method for on-line measurement of abrasion loss of mechanical sealing end face based on technology of digital B-ultrasound imaging |
CN107392364A (en) * | 2017-07-12 | 2017-11-24 | 河海大学 | The short-term load forecasting method of variation mode decomposition and depth belief network |
-
2018
- 2018-07-11 CN CN201810754424.5A patent/CN109099964A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102313578A (en) * | 2011-08-04 | 2012-01-11 | 广州市香港科大霍英东研究院 | Mechanical seal online monitoring system |
CN105115453A (en) * | 2015-04-27 | 2015-12-02 | 南京林业大学 | Device and method for on-line measurement of abrasion loss of mechanical sealing end face based on technology of digital B-ultrasound imaging |
CN107392364A (en) * | 2017-07-12 | 2017-11-24 | 河海大学 | The short-term load forecasting method of variation mode decomposition and depth belief network |
Non-Patent Citations (8)
Title |
---|
万轶: "表面织构与合金化改善密封材料摩擦学性能研究", 《中国博士学位论文全文数据库工程科技Ⅰ辑》 * |
何洋: "《常规兵器电磁兼容性试验》", 1 December 2016, 国防工业出版社 * |
张尔卿: "机械密封端面状态监测及寿命预测关键技术研究", 《中国博士学位论文全文数据库工程科技Ⅱ辑》 * |
张强等: "利用一种模糊神经网络对遥感影像进行分类的研究", 《测绘工程》 * |
李晓晖: "流体动压型机械密封端面状态监测及健康评估", 《中国博士学位论文全文数据库工程科技Ⅱ辑》 * |
田质广等: "模糊B样条神经网络在变压器超高频局部放电模式识别中的应用", 《电子技术应用》 * |
白辰等: "基于模糊神经网络扰动观测器的RLV动态逆再入姿态控制", 《中南大学学报(自然科学版)》 * |
黄伯云等: "《2009年全国青年摩擦学学术会议论文集》", 5 May 2009, 湖南省地质测绘印刷厂 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114644276A (en) * | 2022-04-11 | 2022-06-21 | 伊萨电梯有限公司 | Intelligent elevator control method under mixed scene condition |
CN115754010A (en) * | 2022-11-10 | 2023-03-07 | 四川科力特硬质合金股份有限公司 | Quality control method for sealing end face of mechanical sealing ring |
CN115754010B (en) * | 2022-11-10 | 2024-04-09 | 四川科力特硬质合金股份有限公司 | Quality control method for sealing end face of mechanical sealing ring |
CN116989937A (en) * | 2023-09-25 | 2023-11-03 | 苏州俊煌机械科技有限公司 | Detection method and device for mechanical sealing element |
CN116989937B (en) * | 2023-09-25 | 2023-12-22 | 苏州俊煌机械科技有限公司 | Detection method and device for mechanical sealing element |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Yongbo et al. | Review of local mean decomposition and its application in fault diagnosis of rotating machinery | |
CN109099964A (en) | Mechanical seal end surface state monitoring method | |
Kang et al. | Experimental study on the three-dimensional flow within a compressor cascade with tip clearance: Part I—Velocity and pressure fields | |
Guo et al. | Rolling bearing fault classification based on envelope spectrum and support vector machine | |
Yang et al. | Vibration signal analysis using parameterized time–frequency method for features extraction of varying-speed rotary machinery | |
Bin et al. | Early fault diagnosis of rotating machinery based on wavelet packets—Empirical mode decomposition feature extraction and neural network | |
CN109827777B (en) | Rolling bearing fault prediction method based on partial least square method extreme learning machine | |
Tang et al. | A fault diagnosis method for loose slipper failure of piston pump in construction machinery under changing load | |
CN102033984A (en) | Method for fault diagnosis of rotating mechanical equipment based on interval-type evidence fusion | |
CN107727333A (en) | A kind of diagnostic method for hydraulic cylinder leakage analyzing | |
CN103207070A (en) | Method for diagnosing failure of rotary machine by fusing oil and vibration | |
CN109342046A (en) | A kind of seal assembly floating performance Online Transaction Processing | |
CN114417924B (en) | Rolling bearing fault diagnosis method based on undirected graph adjacent matrix of mixed features | |
Dong et al. | Review of research on intelligent diagnosis of oil transfer pump malfunction | |
Jie et al. | Research on intelligent recognition of axis orbit based on Hu moment invariants and fractal box dimension | |
CN107559228A (en) | The method for being detected based on bispectral data and diagnosing fan trouble | |
Zhao et al. | Intermittent-chaos-and-cepstrum-analysis-based early fault detection on shuttle valve of hydraulic tube tester | |
Wei et al. | Cavitation state recognition for control valve using AlexNet-type neural networks with three-channel images transformed by time series | |
Grasso et al. | A new approach for online health assessment of abrasive waterjet cutting systems | |
Ge et al. | Rough set neural network feature extraction and pattern recognition of shaft orbits based on the zernike moment | |
Ito et al. | An Analysis Of Bottleneck Technology By Using Experiments And CAE-Example Of The Automotive Transaxle Oil Seal Leakage | |
Daraz et al. | Detection and diagnosis of mechanical seal faults in centrifugal pumps based on acoustic measurement | |
Yacout et al. | Monitoring and control of machining process by data mining and pattern recognition | |
CN105205602B (en) | A kind of extension event static classification method towards green products instance analysis | |
Tian et al. | Few-shot condition diagnosis of rolling bearing using adversarial transfer network with class aggregation-guided |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181228 |